CGAL users discussion list

[Monique Teillaud <>]

Hi,

I did not look into the details, but I can see that you are using
Cartesian_d<double> as a kernel, which is NOT recommended.

see
http://www.cgal.org/FAQ.html#inexact_NT
http://www.cgal.org/philosophy.html

best,
--
Monique Teillaud
http://www.inria.fr/sophia/members/Monique.Teillaud/
INRIA Sophia Antipolis - Méditerranée
Institut National de Recherche en Informatique et Automatique

Le 27/03/14 17:21, Ge Zhang a écrit :
> Dear All,
>
> I have figured out a workaround. With a bit experiment, I found that the 
> first 5 points are on the same plane, and the number of simplices becomes 
> incorrect after the 6th point is added. Based on this observation, I added a 
> small (1e-12) perturbation to the coordinates and the number of simplices 
> becomes correct! So my problem is solved. But still, I think this 
> demonstrates that the dD Delaunay package is not robust. It would be good if 
> someone can improve it.
>
> Best,
> Ge
>
>
> From: Ge Zhang
> Sent: Tuesday, March 25, 2014 1:49 PM
> To: 
> ''
> Subject: CGAL Problem
>
> Dear Sir or Madam,
>
> I am not sure if I have made an error or if there is a bug in CGAL dD 
> Delaunay package. I need to do Delaunay triangulation of some 
> three-dimensional points and need to generalize the code to other dimensions 
> later. So I wrote this code:
>
> ///////////////////////////////////////////////////////////////////////////
> double coord[] = {
> 15.987446522913068, 27.202164077804600, -1.7503256142623180,
> 15.783731654803892, 27.268252018459251, -1.7667168014850418,
> 15.366332754158634, 25.364059942735178, -0.12067529139448278,
> 14.798358756496864, 25.645272186454374, -0.24621719891310628,
> 14.416741390968623, 26.534931393517535, -0.90761123478229566,
> 12.653889524424551, 30.356776464885129, 1.6780311801981169,
> 20.141382962264476, 29.666174471723380, 0.81578848925181291,
> 15.255480206458079, 32.065555891106911, 0.14008969527418605,
> 15.856413673274881, 31.537804468497811, -0.50108136475224174,
> 19.537758401022533, 29.790220202824923, -0.21754568768545338,
> 19.418171028905622, 29.790007422401796, -0.38830723957693186,
> 20.081590302704605, 29.854788525235321, 0.73931846754153374,
> 16.096713953744956, 24.603235225447076, 1.9663067370703267,
> 15.057336655250380, 24.912757478035189, 2.7258085800022718,
> 19.717425535978570, 27.248357239016887, 3.5216075631300305,
> 20.340124881568187, 28.287915509305925, 2.9881740318160834,
> 18.440406384399445, 28.756603149763084, 5.7274305773261025,
> 19.732584195169082, 29.437336063389296, 4.1695063361795697,
> 16.615174375063219, 32.278894713132360, 2.6796054812782928,
> 19.411374934889903, 30.874452428442833, 2.6765292926211286,
> 15.527857372851596, 31.342303101257738, 4.4163514428924291,
> 17.970992967584714, 28.796276720304878, 5.9574834972343362,
> 14.323514965294358, 26.455018287900284, 4.7555968341210084,
> 12.748877146187398, 30.055258023558022, 3.4117720135679241,
> };
>
>
> #include <CGAL/Cartesian_d.h>
> #include <CGAL/Kernel_d/Sphere_d.h>
> #include <CGAL/Delaunay_d.h>
> #include <iostream>
> #include <list>
>
> typedef CGAL::Cartesian_d<double> Kernel;
> typedef CGAL::Delaunay_d<Kernel> Delaunay;
> typedef Delaunay::Vertex_handle Vertex_handle;
> typedef Delaunay::Simplex_handle Simplex_handle;
> typedef Delaunay::Point_d Point;
>
>
> int main()
> {
>                  Delaunay De(3);
>                  for(int i=0; i<24; i++)
>                  {
>                                  Point p(3, coord+i*3, coord+i*3+3);
>                                  De.insert(p);
>                  }
>                  std::list<Simplex_handle> simplices = De.all_simplices();
>                  std::cout<<simplices.size();
>                  return 0;
> }
> ///////////////////////////////////////////////////////////////////////////
>
> When I run this program on a Windows Server 2012 machine using Visual Studio 
> 2012 compiler, the output I get is:
> 3
>
> The problem is, if one plot the coordinates using MATLAB, one can clearly see 
> that those points form a polygon and the Delaunay triangulation should 
> consist of way more than 3 simplices. But why does the program only finds 3 
> simplices?
>
> Best,
> Ge Zhang